Fig. 4. ed acts in the canonic N pathway, before the cytoplasmic release of N^ICD. (A,B) Removal of most microchaetae by expression of hs-N^icd in 0-8 hours-old pupae (A) is not rescued by UAS-ed^ECD driven by sca-Gal4 (B). The control shown in Fig. 3G indicates that UAS-ed^ECD driven by sca-Gal4 was active during microchaetae determination, as it increased microchaetae density. The transformation of macrochaetae to double sockets, owing to the excess of N signaling during differentiation, is clearly visible in some cases (arrowheads). (C,D) In a similar experiment, the weak removal of microchaetae by hs-N^ECN (C) is not rescued by UAS-ed^ECD (D). (E,F) The low density of microchaetae typical of N^mcd1/+ (E) is not rescued by decreasing ed function (F, ed^1X5/ed^slH8 combination). (G,H) The absence of macrochaetae characteristic of Ax^M1/+ (G) is not rescued by ed^1X5/ed^slH8 (H).